New Istanbul Airport is constructed over an area of 76.5 million square metres on a former open-cast coalfield located on the European side of Istanbul, 35km northwest of the city centre towards the Black Sea coast. The airport hosted 23.4 million passengers in 2020 despite the serious decrease in the number of passengers at all airports worldwide due to the Covid-19 pandemic. It is planned to offer flights to more than 350 destinations globally.
The airport is currently connected to the city with a fleet of buses, but the metro will make its way to the new airport by 29 October 2021, the date of the foundation of the Turkish Republic. There are two lines with two tubes each, one with a length of 37.5km from Gayrettepe to the airport, and another with a length of 35km from Halkali to the airport. The construction of the second line from Halkali to the airport is underway.
The first line, from Gayrettepe to the airport, has nine stations and a total (twin tube) length of 75km (of which 69km was bored by TBM). In order to excavate the long TBM tunnel drives, 10 EPB TBMs were used (6.56m diameter) from four different TBM suppliers.
The geology along the TBM tunnel routes is complex, with several transition zones, shear zones and faults. The line is planned for 150,000 passengers per hour, with travel times between the city centre and the airport to take around 24 minutes.
GEOLOGY
Prevailing geological strata along the tunnel alignment include Quaternary, Tertiary and Paleozoic-aged Trakya Formations; these comprise interbedded sandstone, mudstone, siltstone of different degrees of weathering, and overlying Tertiary-aged Danisment Formation consisting of claystone, silty, and sand with weak sandstone bands. Because Danisment Formation bears methane, it necessitated a detailed risk analysis prior to excavation. In the light of experiences gained from methane/dust explosions, conveyor belt fires in Turkish coal mines, and fires generated thereafter, guiding principles of managing the risk of tunnel fires during TBM excavations are developed and discussed in a recent paper published in Tunnelling and Underground Space Technology (Bilgin, Balci, Arslanbas, 2021). Several faults, weak zones, transition zones and a few lakes left after mining activities along the alignment caused tremendous difficulties during tunnel excavations (Arslanbas and Bilgin, 2020).
LESSONS LEARNT Appropriate EPB face pressure decreases the risk of TBM blockages and jamming
The Istanbul Airport project demonstrated that one of the key factors in the success of EPB TBM drives in complex geology is the proper use of EPB face pressure. The concept given by Anagnostou and Kovari (1996) for the calculation of EPB face pressure was used for both lines in the Istanbul Airport Metro Project tunnels. However, previous studies showed that none of the face pressure calculation methodologies gave the correct EPB pressure in Istanbul’s complex geology.
An EPB report was prepared for each kilometre of tunnel length and the values later compared to the actual values. The next EPB report for the upcoming kilometre was prepared by calibrating the calculated values taking into consideration the previously obtained actual values and using the accumulated EPB data from the similar ground conditions.
Two of the TBMs (the one in the first tube being 30m in front of the TBM in the second tube) jammed/ blocked consecutively within faults F1, F2 and F3 due to the complex weak zones with loose and open crushed areas between rings 2656/2856 (chainages 5 + 260/4 + 960 km). These zones demonstrate a complex geometry and structure, and are filled with materials of very low strength characteristics.
A typical view of one of the complex weakness zones (Fault 1) is seen in Figure 1. In this area, the TBM in the first tube jammed three times but the one in the second tube jammed only once. Initially, remedial works for the first blockage involved injecting bentonite and polymer MasterRoc SLP 2 liquid polymer at a pressure of 2bar through 16 holes opened around the tail shield.
After various unsuccessful attempts, it was decided to use five additional jacks installed between the segments and the tail shield, achieving an extra 15,000kN thrust force. A further additional thrust of 10,000kN was also obtained using 16 passive articulation cylinders.
As seen from Figure 2, the TBM jammed at points 1, 2 and 3. In each case, additional jacks were used to free the TBM which moved 4.5m, 4m and 0.8m respectively before the rescue operation was completed. However, excavation time increased from 30 to 60min/ring, rising to 752min/ring and decreasing thereafter to 510min/ring and 120min/ring during salvage operations.
To compare TBM performance between the first and second tubes, approximate critical ring numbers were selected based on Figures 2 and 3, showing variations in EPB pressure with ring numbers in tube 1 and tube 2. As seen from Figure 2, EPB pressure changed between 0.45 and 0.94 within points A, B, and C, with a mean value of 0.7 between ring numbers 2410/2618 for tube 1.
After the first blockage, EPB pressure fluctuated between 0.47 and 4bar between rings 2618-2880, with a mean value of 1.1bar. Figure 3 shows the variation of EPB pressure with ring numbers for tube 2. Referring to this figure, between chainages 2410-2518 (points A, B), 2518-2618 (points B, C) and 2618-2880 (points C, D) EPB pressure increased consecutively, having mean values of 1.2, 1.64, and 2.1bar.
The most notable difference in the performance of the two machines was that the TBM jammed three times in tube 1 and only once in tube 2. The mean EPB pressure applied in the second tube within the studied chainages was 1.65bar, which was almost twice the one used in the first tube (0.83). It is concluded that higher EPB pressures helped to stop the collapse of loose materials within the complex weakness zones. However, it is questionable how far away a face pressure can affect the collapse of loose weak zones upon the tail shield which may depend on the dimensions and the inclination of weak zones. It is believed that this is an open area for further research, probably using numerical analysis methods. The reader is advised to consult the paper by Acun, Bilgin, and Erboylu, 2012 for more information on the effect of proper EPB pressure as a mitigation factor to prevent TBM blockages.
PREPARE FOR HYPERBARIC INTERVENTION
Figure 4 shows a typical geologic cross section with several transition zones between 11,700m and 12,600m. One of the main problems of TBM excavation in the transition zones given in Figure 4 was a high cutter consumption. This was especially the case when transitioning from soft ground to hard rock formations, where the cutters were destroyed and the cutterhead also damaged. Excessive disc consumption and cutterhead damage also necessitated hyperbaric interventions. The abrasiveness of the material excavated also contributed to this undesirable situation. Therefore, under high EPB pressures, hyperbaric intervention was inevitable at chainage 11+700km (Erboylu et al. 2020).
Figure 5 shows a diver working under a 3.2bar hyperbaric intervention. Each diver was allowed to work one hour in hyperbaric conditions and, as a group of four, divers were able to change 18 disc cutters in one week. This underlined the fact that TBMs should have a suitable man-lock for hyperbaric interventions and the contractor must be ready to take this action.
SINKHOLES MAY OCCUR EVEN WITH 80m OVERBURDEN
Transition zones between two different geologic formations have generally weak interfaces and sometimes cause serious problems in tunnel excavation. The common problems include an inability to maintain face pressure, ground loss and sinkholes. In our case, a sinkhole occurred between chainages 12+130/12+120km at the transition zone between weathered sandstone and massive sandstone, and at a maximum EPB pressure of 3.8bar (Figure 4). The sinkhole was 80m above the crown of the tunnel and required 800m3 of concrete to fill it.
Different cases experienced in Turkey have shown that probe drilling and umbrella arch techniques can help to avoid unwanted scenarios, such as TBM squeezing/ blocking or sinkholes in complex geology (Bilgin, Copur, Balci, 2014). However, the examples given above showed clearly that using the proper EPB face pressure and controlling ground losses may also be used as mitigation factors in adverse geological conditions.
TRANSITION ZONES DECREASE MACHINE USAGE TIME AND INCREASE CUTTER CONSUMPTION
The most dramatic effect of the transition zone was on machine usage time since excessive mud coming from the belt conveyor caused long delays in cleaning the tunnel. The other negative effect of excavation in transition zones was on daily advance rates (Figure 7).
Time spent on inspecting the cutterhead and on changing the cutters increased by a factor of up to 3.5. Especially when the TBM transitioned from soft ground to hard rock formations, the cutters were wrecked and the cutterhead was also damaged. This made hyperbaric intervention inevitable.
To better cope with transition zones, it is strictly recommended that the interpretation of geology must be of prime concern. The geotechnical properties of materials encountered in transition zones should be well studied for ground conditioning. In addition, the tunnel management must be ready to cope with difficulties that will be created by excavating in transition zones (Bilgin and Acun., S, 2021).
RISK ANALYSIS AND SECURITY PROGRAMMES NECESSARY IN METHANE-BEARING STRATA
The airport construction area was on a former coalfield – Oligocene-aged coal seams comprising upper, middle and lower seams. The mine is currently abandoned and one of the biggest airports in the world is being constructed on it. The possibility of encountering methane in the excavated areas between chainages 24+375km and 24+575km was a big concern since previously, a methane explosion inside the excavation chamber of a TBM boring the Selimpasa wastewater tunnel, Turkey resulted in several workers being injured (Bilgin, Copur and Balci, 2014).
In the Selimpasa tunnel, methane from a fault zone accumulated in the TBM’s excavation chamber and an explosion occurred due to sparking, caused most probably by contact between the screw conveyor and its casing. Methane (CH4) is a colourless, odourless and non-noxious gas. Since the density of methane (0.716kg/ m3) is lower than the density of air (1.293kg/m3), it accumulates in the upper part of a tunnel or gallery. When it mixes with air, it becomes flammable and explosive, creating a danger. When the concentration of methane within the air is between 5% (lower limit) and 15% (upper limit), it becomes explosive.
Within chainages 24+375km and 24+575km (where the coal seams were excavated), the following safety programme was initiated:
a) Smoking was strictly prohibited in the tunnel.
b) Workers were reminded never to forget that the methane content in the pressure chamber of an EPBTBM is always higher than the methane content in the tunnel, since methane in the air inside the tunnel is always diluted by tunnel ventilation. Also, the oxygen found in the foam is a very favorable medium to create a methane explosion.
c) When an air sample indicates a lower explosive limit (LEL) of 5% or more, the air supply should be increased with continuous gas control; a minimum number of workers should be employed around the screw conveyor since flame and muck get around the tunnel through the muck chamber of the screw conveyor as observed at the Selimpasa Tunnel. A coal-dust explosion is worse than a methane explosion and maximum care should be taken to avoid coal-dust accumulations around the belt conveyor and TBM area. Fire-resistant conveyor belts must be used; welding or cutting operations are prohibited next to belt conveyors and hydraulic power lines. As methane is soluble in water, it migrates with groundwater and reaches the tunnel wherever possible so gas bubbling should be checked throughout the tunnel. Apart from gas sensors on the TBM, extra gas sensors should be located at critical points. Methane generated within the pressure chamber must be controlled with a special tube. When an air sample indicates 5% LEL or more, the air supply should be increased.
CONCLUSIONS
New Istanbul Airport is constructed on a former coal mine and in a complex geology. The following lessons may be drawn from EPB-TBM drives in such geological conditions:
? Using appropriate EPB face pressures decreases the risk of TBM blockages.
? TBMs should have a suitable man-lock for hyperbaric interventions and contractors must be prepared for achieving this.
? Transition zones decrease machine usage time and increase cutter consumption, meaning that project schedules must be prepared according to the prevailing conditions.
? The geotechnical properties of the ground within transition zones should be well studied to ensure correct ground conditioning.
? A sinkhole can occur even at 80m of overburden.
? The tunnel management must be prepared to cope with the difficulties associated with excavating in transition zones.